auto_awesome Semantic Abstract Library

Mapping the Abstract World

Every icon is a geometric primitive mapped to a human concept. Discover the visual DNA of language through conceptual exploration.

search
811
Icons
25k
Goal
SVG
Vector

DATA BLOCKED: Open with VS Code "Live Server" to load icons.

Browse by Essence

Abstract

Conceptual ideas and intangible forms.

System

Interface elements and structural primitives.

Motion

Dynamic energy and directional flow.

Feeling

Human emotion and inner states.

Our Methodology

Where Language Meets Geometry

We map semantic nodes — words, concepts, emotions — onto a coordinate system of geometric primitives. Each icon is generated following strict protocols for stroke weight, curvature, and visual hierarchy.

text_fields

Semantic Input

Linguistic data is parsed into primary semantic dimensions.

polyline

Geometric Mapping

Attributes like curvature, symmetry, and weight are assigned algorithmically.

auto_awesome

AI Generation

Icons rendered via Gemini 2.0 Flash, validated against our geometric protocol.

Concept STABILITY
Mapping sync_alt
Form SQUARE

COORD: 45.2 // SEMANTIC_TONE: STABLE

Language Directory

Customize

Visual DNA

Conceptual Mapping

Related Icons

LLOS.ai Asset Open-License
ballot 30-Second Poll

Help Shape This Library

Your feedback directly influences which icons we design next. Five quick questions, meaningful impact.

Frequently Asked Questions

Everything you need to know about the Visual Explorer.

Visual Explorer is a comprehensive icon library that maps abstract concepts to geometric primitives. Each icon is designed following strict geometric protocols with consistent stroke weights and primitives, creating a unified visual language system.

Use the search bar at the top to find icons by name, category, or keywords. As you type, suggestions appear to help you discover related concepts. The search supports partial matches and semantic associations through our expansion engine.

Icons are available as SVG (Scalable Vector Graphics) and PNG files. SVGs are resolution-independent and fully customizable with CSS or JavaScript. PNGs are exported at your chosen size with background and stroke colors baked in. Each icon uses a standard 24x24 viewBox. You can also copy raw SVG code or JSX directly from the detail panel.

Yes, there is a limit of 10 downloads per session, tracked via your browser's local storage. This helps us manage resources while allowing exploration. The counter resets if you clear browser data.

Please refer to our licensing terms for commercial use. The icons are designed for educational and creative purposes within the LLOS.ai ecosystem. For commercial licensing inquiries, contact our team.

Icons are organized into primary categories (Abstract, System) plus rich keyword tags for semantic searching. Each icon has multiple tags for granular discovery. The sidebar filters let you narrow by category, and quick-tags surface popular concepts.

Our icons are designed with a unique geometric protocol that emphasizes semantic meaning over decoration. Each icon represents the conceptual essence of a word, creating a visual language that bridges abstract thinking and geometric form.

We're actively generating new icons using AI-assisted design. Our goal is 25,000 unique icons. Updates are released in batches as they pass our quality and consistency standards.

Yes! You can customize icons right here in the browser before downloading. Click any icon to open the detail panel, then use the Customize section to change background color, stroke color, and size. Downloads include your customizations. After downloading, SVG files remain fully editable in Figma, Illustrator, Inkscape, or with CSS in web projects.

The accent class highlights 1-2 key elements within an icon using the saffron gold color, creating visual hierarchy and emphasis that guides the eye to the most important part of the symbol.

Developers can access raw JSON data with all icon metadata, SVG paths, categories, and keywords. This enables programmatic search, dynamic rendering, and seamless integration into any web or mobile application. Use "Copy JSX" for React projects.

Visual Explorer works on all modern browsers including Chrome, Firefox, Safari, and Edge. Mobile browsers are fully supported.

We welcome feedback! Report issues or suggest new icons through our contact form. Community input helps us prioritize which concepts to visualize next.

Icons are generated using Google's Gemini 2.0 Flash AI model, following our proprietary geometric design protocol. Each icon is validated for consistency against strict visual language standards.

We're exploring community contributions for future phases. All contributions must follow our geometric protocol and pass our quality review. Reach out if interested.

Click any icon to open the detail panel. Below the icon preview you'll find the Customize section with three controls: Background (pick from presets or use the color picker), Stroke Color (change the line color of the icon), and Size (use the slider or quick presets from 24px to 512px). All changes preview live and are baked into your download.

Yes! Use the PNG button in the detail panel to export a raster image at your selected size. The PNG includes your chosen background color and stroke color baked in. For scalable use, SVG remains the recommended format.

The LLOS Visual Grammar is the design constitution behind every icon. It defines 5 Laws (accent = meaning token, one concept = one gesture, semantic position is sacred, shared primitives only, encode relationships not objects), 7 POS visual modes (Noun, Verb, Adjective, etc.), and the saffron accent principle where gold highlights carry the semantic payload of each icon.